Editorial on “Can CT-PET and endoscopic assessment post-neoadjuvant chemoradiotherapy predict residual disease in esophageal cancer”

Neoadjuvant chemoradiation (nCRT) followed by surgery is the most common treatment modality for resectable esophageal cancer nowadays (1). Esophagectomy and gastrointestinal continuity reconstruction are extensive and challenging procedures (2). Some studies had shown that patients achieved pathologic complete response (pCR) may not benefit from subsequent surgery (3,4). However, the assumption of sparing the surgery is based on the pathologically disease free of patients. The main question arise that, can the clinical complete response (cCR) assessed by positron emission tomography (PET)/computed tomography (CT) and endoscope truly represent the microscopic complete response on pathologic examination?

In the article “Can CT-PET and endoscopic assessment post-neoadjuvant chemoradiotherapy predict residual disease in esophageal cancer” recently published in Annals of Surgery (5), Heneghan et al. tried to answer the question. The authors performed a retrospective analysis on consecutive patients with stage I–III locally advanced esophageal cancer (LAEC) between 2006 and 2014. All 138 patients [103 adenocarcinomas and 35 squamous cell carcinomas (SCC)] had nCRT followed by surgery. Both endoscopy and F-18-fluorodeoxyglucose (¹⁸F-FDG) PET/CT evaluations were performed at the initial staging and at restaging 4–6 weeks after nCRT. Authors investigated the ability of the restaging PET/CT and endoscope in identification of patients with pCR.

Complete metabolic response (cMR), defined by SUVmax <4 and no nodal uptake on PET/CT, showed poor sensitivity (55.9%) and poor positive predictive value (PPV, 30.2%) for pCR. Among the 63 patients with cMR, only 30% had actually achieved pCR, while many of them had minimal residual disease (32%) or nodal positive disease (25%). The change in tumor SUVmax (%ΔSUVmax, 63.3%±24.7%) was a significant predictor of pCR in univariate analysis but was not significant in multivariate analysis. Complete response on endoscopy (cER) also had poor sensitivity (40.7%) and PPV (24.4%) for pCR. Although endoscopic ultrasound and systematic biopsy were not performed in this study, the low sensitivity of endoscopy in detection of pCR is consistent with other studies (6-8). The complete responder groups defined by PET/CT and endoscope seem not consistent with each other (Spearman correlation coefficient =0.07, P=0.48). This discrepancy between the two modalities may be explained by the basic difference between anatomic and molecular imaging. Unlike the direct visualization of morphologic changes of esophageal mucosa on endoscope, the metabolic changes on PET may precede detectable morphologic changes on endoscope (9). However, both modalities are suboptimal in detection of pCR when the mucosal changes or glycolytic activity of minimal tumor burden is beyond their resolutions. The cCR, defined as combination of both cMR and cER, was limited by low sensitivity of both modalities and had the worst sensitivity (32.4%) and poor PPV (35.5%) for pCR. No matter using cMR, cER or cCR, the sensitivity and PPV for pCR were worse on patients with adenocarcinoma than SCC. In multivariate analysis, only histological subtype and lymph node status were significant predictors of pCR. None of the imaging assessments was independent predictor of pCR. However, both cMR and cCR demonstrated prognostic significance in survival cox regression analysis. Of interest, in patients with pCR, those with cMR had additional survival benefit. The importance of PET/CT in post-nCRT risk stratification had been well known (10,11).

The intention to avoid surgery for LAEC is not only because of the reduction of surgical risks but also the improved quality of life. Reliable assessment tool that could identify pCR is the assumption of this personalized strategy for LAEC. Although the definition of cMR using SUVmax <4 is debatable, the findings in this study were consistent with others. The accuracy of PET/CT to predict pCR was suboptimal and could not justify the omission of esophagectomy (12-15). The results were not surprised and showed consistency with our clinic experience. The resolution of PET/CT and endoscope cannot detect residual tumor cells or micrometastases and thus has falsely classified patients as complete resolution. This was manifested as low specificity of PET/CT for pCR in Heneghan’s work (5) and (12). Interestingly, in Heneghan et al., the accuracy of PET/CT was limited in sensitivity as well as in specificity. Low sensitivity and thus high false negative rate of cMR meant that patients with pCR might have not been revealed as cMR. Despite the adequate 4 to 6 weeks waiting time before restaging PET/CT, FDG uptake caused by radiation-induced inflammation could possibly have not subsided yet. Furthermore, the difficulty in differentiating gastric uptake from residual adenocarcinoma located at esophagogastric junction could have contributed to the falsely residual FDG uptake. It is not uncommon that SUVmax could be higher than four in both scenarios. The definition of cMR may be further adjusted in future studies.

The %ΔSUVmax was a significant predictor of pCR in univariate analysis [odds ratio 1.03 (1.01–1.06), P<0.01]. Although it was not significant in multivariate analysis, it is worth of noticing that PET parameters quantifying the changes between pre-nCRT and post-nCRT might be more predictive than those based solely on post-nCRT PET. Radiomics, which uses computerized tools to extract a large number of image features, is an emerging quantitative imaging biomarker in oncology (16). Several studies applied radiomics for esophageal cancer, especially for prediction of treatment response and prognosis (17,18). PET texture features outperformed SUVmax in identification of partial or complete responder to nCRT, defined by RECIST criteria (19), with sensitivity of each feature ranged from 76% to 92%. We found features derived from PET intensity and texture had equal or better accuracy than %ΔSUVmax for the prediction of pathologic tumor response to nCRT (18). The post-nCRT PET of responder tended to be more homogenous on texture features. For better prediction, the radiomics features were combined to construct a multivariate regression model or machine learning models (20-22). We constructed a support vector machine model with radiomics features and clinical parameters, which achieved a sensitivity higher than 90% for predicting partial or complete pathologic response to nCRT (20). It would be more challenging when the task is restricted to predict complete pathologic response only. Nevertheless, Tixier et al. showed that some texture features can identify complete pathologic response better than SUV-based parameters (17). Recently, Desbordes et al. built a random forest classifier for cCR with sensitivity 82%±9% and specificity 91%±12% (21). For prediction of pCR, van Rossum et al. found that the incorporation of texture and geometry features could improve the performance of prediction models with clinical factors and conventional PET parameters (23). Though many issues in radiomics are still needed to be investigated (24), it is one of the most promising tools in the era of precision medicine considering its potential and relative low additional cost.

¹⁸F-FDG PET/CT has been a valuable imaging tool in the prognosis and response evaluation of esophageal cancer. However, the complete resolution of tumor on PET/CT after nCRT is of prognostic significance but is not sufficiently reliable to serve as the decision maker for avoiding complex surgery for LAEC. Heneghan’s recent work highlighted the low sensitivity and positive predict value of PET/CT for identifying pathologic complete resolution after nCRT. Other prediction tool or quantification techniques like radiomics should be investigated before the application of surgery-as-needed strategy for esophageal cancer.

Acknowledgements

Funding: This work was supported in part through the NIH/NCI Grant R01CA172638 and the NIH/NCI Cancer Center Support Grant P30 CA008748.

Footnote

Conflicts of Interest: The authors have no conflicts of interest to declare.

References

1. van Hagen P, Hulshof MC, van Lanschot JJ, et al. Preoperative chemoradiotherapy for esophageal or junctional cancer. N Engl J Med 2012;366:2074-84. [Crossref] [PubMed]
2. Paul S, Altorki N. Outcomes in the management of esophageal cancer. J Surg Oncol 2014;110:599-610. [Crossref] [PubMed]
3. Bedenne L, Michel P, Bouche O, et al. Chemoradiation followed by surgery compared with chemoradiation alone in squamous cancer of the esophagus: FFCD 9102. J Clin Oncol 2007;25:1160-8. [Crossref] [PubMed]
4. Stahl M, Stuschke M, Lehmann N, et al. Chemoradiation with and without surgery in patients with locally advanced squamous cell carcinoma of the esophagus. J Clin Oncol 2005;23:2310-7. [Crossref] [PubMed]
5. Heneghan HM, Donohoe C, Elliot J, et al. Can CT-PET and Endoscopic Assessment Post-Neoadjuvant Chemoradiotherapy Predict Residual Disease in Esophageal Cancer? Ann Surg 2016;264:831-8. [Crossref] [PubMed]
6. van Rossum PS, Goense L, Meziani J, et al. Endoscopic biopsy and EUS for the detection of pathologic complete response after neoadjuvant chemoradiotherapy in esophageal cancer: a systematic review and meta-analysis. Gastrointest Endosc 2016;83:866-79. [Crossref] [PubMed]
7. Schneider PM, Metzger R, Schaefer H, et al. Response evaluation by endoscopy, rebiopsy, and endoscopic ultrasound does not accurately predict histopathologic regression after neoadjuvant chemoradiation for esophageal cancer. Ann Surg 2008;248:902-8. [Crossref] [PubMed]
8. Griffin JM, Reed CE, Denlinger CE. Utility of restaging endoscopic ultrasound after neoadjuvant therapy for esophageal cancer. Ann Thorac Surg 2012;93:1855-9; discussion 1860.
9. Yanagawa M, Tatsumi M, Miyata H, et al. Evaluation of response to neoadjuvant chemotherapy for esophageal cancer: PET response criteria in solid tumors versus response evaluation criteria in solid tumors. J Nucl Med 2012;53:872-80. [Crossref] [PubMed]
10. Miyata H, Yamasaki M, Takahashi T, et al. Determinants of response to neoadjuvant chemotherapy for esophageal cancer using 18F-fluorodeoxiglucose positron emission tomography (18F-FDG-PET). Ann Surg Oncol 2014;21:575-82. [Crossref] [PubMed]
11. Cervino AR, Pomerri F, Alfieri R, et al. 18F-fluorodeoxyglucose PET/computed tomography and risk stratification after neoadjuvant treatment in esophageal cancer patients. Nucl Med Commun 2014;35:160-8. [Crossref] [PubMed]
12. Cheedella NK, Suzuki A, Xiao L, et al. Association between clinical complete response and pathological complete response after preoperative chemoradiation in patients with gastroesophageal cancer: analysis in a large cohort. Ann Oncol 2013;24:1262-6. [Crossref] [PubMed]
13. Molena D, Sun HH, Badr AS, et al. Clinical tools do not predict pathological complete response in patients with esophageal squamous cell cancer treated with definitive chemoradiotherapy. Dis Esophagus 2014;27:355-9. [Crossref] [PubMed]
14. Kwee RM. Prediction of tumor response to neoadjuvant therapy in patients with esophageal cancer with use of 18F FDG PET: a systematic review. Radiology 2010;254:707-17. [Crossref] [PubMed]
15. Ngamruengphong S, Sharma VK, Nguyen B, et al. Assessment of response to neoadjuvant therapy in esophageal cancer: an updated systematic review of diagnostic accuracy of endoscopic ultrasonography and fluorodeoxyglucose positron emission tomography. Dis Esophagus 2010;23:216-31. [Crossref] [PubMed]
16. Gillies RJ, Kinahan PE, Hricak H. Radiomics: Images Are More than Pictures, They Are Data. Radiology 2016;278:563-77. [Crossref] [PubMed]
17. Tixier F, Le Rest CC, Hatt M, et al. Intratumor heterogeneity characterized by textural features on baseline 18F-FDG PET images predicts response to concomitant radiochemotherapy in esophageal cancer. J Nucl Med 2011;52:369-78. [Crossref] [PubMed]
18. Tan S, Kligerman S, Chen W, et al. Spatial-temporal [(1)(8)F]FDG-PET features for predicting pathologic response of esophageal cancer to neoadjuvant chemoradiation therapy. Int J Radiat Oncol Biol Phys 2013;85:1375-82. [Crossref] [PubMed]
19. Therasse P, Arbuck SG, Eisenhauer EA, et al. New guidelines to evaluate the response to treatment in solid tumors. European Organization for Research and Treatment of Cancer, National Cancer Institute of the United States, National Cancer Institute of Canada. J Natl Cancer Inst 2000;92:205-16. [Crossref] [PubMed]
20. Zhang H, Tan S, Chen W, et al. Modeling Pathologic Response of Esophageal Cancer to Chemoradiation Therapy Using Spatial-Temporal (18)F-FDG PET Features, Clinical Parameters, and Demographics. Int J Radiat Oncol Biol Phys 2014;88:195-203. [Crossref] [PubMed]
21. Desbordes P, Ruan S, Modzelewski R, et al. Predictive value of initial FDG-PET features for treatment response and survival in esophageal cancer patients treated with chemo-radiation therapy using a random forest classifier. PLoS One 2017;12:e0173208. [Crossref] [PubMed]
22. Beukinga RJ, Hulshoff JB, van Dijk LV, et al. Predicting Response to Neoadjuvant Chemoradiotherapy in Esophageal Cancer with Textural Features Derived from Pretreatment 18F-FDG PET/CT Imaging. J Nucl Med 2017;58:723-9. [Crossref] [PubMed]
23. van Rossum PS, Fried DV, Zhang L, et al. The Incremental Value of Subjective and Quantitative Assessment of 18F-FDG PET for the Prediction of Pathologic Complete Response to Preoperative Chemoradiotherapy in Esophageal Cancer. J Nucl Med 2016;57:691-700. [Crossref] [PubMed]
24. Hatt M, Tixier F, Cheze Le Rest C, et al. Robustness of intratumour (1)(8)F-FDG PET uptake heterogeneity quantification for therapy response prediction in oesophageal carcinoma. Eur J Nucl Med Mol Imaging 2013;40:1662-71. [Crossref] [PubMed]